Java:If vs. Switch

Question

我有一段代码与a)我用b代替纯粹的易读性...

一个)

if ( WORD[ INDEX ] == 'A' ) branch = BRANCH.A;
/* B through to Y */
if ( WORD[ INDEX ] == 'Z' ) branch = BRANCH.Z;

b)

switch ( WORD[ INDEX ] ) {
    case 'A' : branch = BRANCH.A; break;
    /* B through to Y */
    case 'Z' : branch = BRANCH.Z; break;
}

...开关版本是否会级联所有排列或跳转到案例？

编辑:

下面的一些答案涉及上述方法的替代方法.
我已经包含以下内容以提供其使用的上下文.

我问的问题,上面的问题,是因为增加词的经验改进的速度.

这不是生产代码,而是作为PoC快速入侵.

以下似乎是对思想实验失败的证实.
我可能需要比我目前使用的词语更大的词汇.
失败的原因是我没有考虑仍需要内存的空引用. (doh!)

public class Dictionary {
    private static Dictionary ROOT;
    private boolean terminus;
    private Dictionary A, B, C, D, E, F, G, H, I, J, K, L, M, N, O, P, Q, R, S, T, U, V, W, X, Y, Z;
    private static Dictionary instantiate( final Dictionary DICTIONARY ) {
        return ( DICTIONARY == null ) ? new Dictionary() : DICTIONARY;
    }
    private Dictionary() {
        this.terminus = false;
        this.A = this.B = this.C = this.D = this.E = this.F = this.G = this.H = this.I = this.J = this.K = this.L = this.M = this.N = this.O = this.P = this.Q = this.R = this.S = this.T = this.U = this.V = this.W = this.X = this.Y = this.Z = null;
    }
    public static void add( final String...STRINGS ) {
        Dictionary.ROOT = Dictionary.instantiate( Dictionary.ROOT );
        for ( final String STRING : STRINGS ) Dictionary.add( STRING.toUpperCase().toCharArray(), Dictionary.ROOT , 0, STRING.length() - 1 );
    }
    private static void add( final char[] WORD, final Dictionary BRANCH, final int INDEX, final int INDEX_LIMIT ) {
        Dictionary branch = null;
        switch ( WORD[ INDEX ] ) {
        case 'A' : branch = BRANCH.A = Dictionary.instantiate( BRANCH.A ); break;
        case 'B' : branch = BRANCH.B = Dictionary.instantiate( BRANCH.B ); break;
        case 'C' : branch = BRANCH.C = Dictionary.instantiate( BRANCH.C ); break;
        case 'D' : branch = BRANCH.D = Dictionary.instantiate( BRANCH.D ); break;
        case 'E' : branch = BRANCH.E = Dictionary.instantiate( BRANCH.E ); break;
        case 'F' : branch = BRANCH.F = Dictionary.instantiate( BRANCH.F ); break;
        case 'G' : branch = BRANCH.G = Dictionary.instantiate( BRANCH.G ); break;
        case 'H' : branch = BRANCH.H = Dictionary.instantiate( BRANCH.H ); break;
        case 'I' : branch = BRANCH.I = Dictionary.instantiate( BRANCH.I ); break;
        case 'J' : branch = BRANCH.J = Dictionary.instantiate( BRANCH.J ); break;
        case 'K' : branch = BRANCH.K = Dictionary.instantiate( BRANCH.K ); break;
        case 'L' : branch = BRANCH.L = Dictionary.instantiate( BRANCH.L ); break;
        case 'M' : branch = BRANCH.M = Dictionary.instantiate( BRANCH.M ); break;
        case 'N' : branch = BRANCH.N = Dictionary.instantiate( BRANCH.N ); break;
        case 'O' : branch = BRANCH.O = Dictionary.instantiate( BRANCH.O ); break;
        case 'P' : branch = BRANCH.P = Dictionary.instantiate( BRANCH.P ); break;
        case 'Q' : branch = BRANCH.Q = Dictionary.instantiate( BRANCH.Q ); break;
        case 'R' : branch = BRANCH.R = Dictionary.instantiate( BRANCH.R ); break;
        case 'S' : branch = BRANCH.S = Dictionary.instantiate( BRANCH.S ); break;
        case 'T' : branch = BRANCH.T = Dictionary.instantiate( BRANCH.T ); break;
        case 'U' : branch = BRANCH.U = Dictionary.instantiate( BRANCH.U ); break;
        case 'V' : branch = BRANCH.V = Dictionary.instantiate( BRANCH.V ); break;
        case 'W' : branch = BRANCH.W = Dictionary.instantiate( BRANCH.W ); break;
        case 'X' : branch = BRANCH.X = Dictionary.instantiate( BRANCH.X ); break;
        case 'Y' : branch = BRANCH.Y = Dictionary.instantiate( BRANCH.Y ); break;
        case 'Z' : branch = BRANCH.Z = Dictionary.instantiate( BRANCH.Z ); break;
        }   
        if ( INDEX == INDEX_LIMIT ) branch.terminus = true;
        else Dictionary.add( WORD, branch, INDEX + 1, INDEX_LIMIT );
    }
    public static boolean is( final String STRING ) {
        Dictionary.ROOT = Dictionary.instantiate( Dictionary.ROOT );
        return Dictionary.is( STRING.toUpperCase().toCharArray(), Dictionary.ROOT, 0, STRING.length() - 1 );
    }
    private static boolean is( final char[] WORD, final Dictionary BRANCH, final int INDEX, final int INDEX_LIMIT ) {
        Dictionary branch = null;
        switch ( WORD[ INDEX ] ) {
        case 'A' : branch = BRANCH.A; break;
        case 'B' : branch = BRANCH.B; break;
        case 'C' : branch = BRANCH.C; break;
        case 'D' : branch = BRANCH.D; break;
        case 'E' : branch = BRANCH.E; break;
        case 'F' : branch = BRANCH.F; break;
        case 'G' : branch = BRANCH.G; break;
        case 'H' : branch = BRANCH.H; break;
        case 'I' : branch = BRANCH.I; break;
        case 'J' : branch = BRANCH.J; break;
        case 'K' : branch = BRANCH.K; break;
        case 'L' : branch = BRANCH.L; break;
        case 'M' : branch = BRANCH.M; break;
        case 'N' : branch = BRANCH.N; break;
        case 'O' : branch = BRANCH.O; break;
        case 'P' : branch = BRANCH.P; break;
        case 'Q' : branch = BRANCH.Q; break;
        case 'R' : branch = BRANCH.R; break;
        case 'S' : branch = BRANCH.S; break;
        case 'T' : branch = BRANCH.T; break;
        case 'U' : branch = BRANCH.U; break;
        case 'V' : branch = BRANCH.V; break;
        case 'W' : branch = BRANCH.W; break;
        case 'X' : branch = BRANCH.X; break;
        case 'Y' : branch = BRANCH.Y; break;
        case 'Z' : branch = BRANCH.Z; break;
        }
        if ( branch == null ) return false;
        if ( INDEX == INDEX_LIMIT ) return branch.terminus;
        else return Dictionary.is( WORD, branch, INDEX + 1, INDEX_LIMIT );
    }
}

Answer 1

不要担心性能; 使用最能表达您正在做的事情的语法.只有在你(a)表现出性能不足之后; (b)将其本地化为相关例程,只有这样你才应该担心表现.对于我的钱,案例语法在这里更合适.

Answer 2

在字节码中有两种形式的开关:tableswitch和lookupswitch.一个假设密集的密钥,另一个密集.请参阅JVM规范中的编译开关说明.对于枚举,找到序数,然后代码继续int.我不完全知道如何提出的switch关于String在JDK7的小功能将得到实施.

但是,大量使用的代码通常在任何合理的JVM中编译.优化者并非完全愚蠢.不要担心,并按照通常的启发式方法进行优化.

Answer 3

看起来你已经枚举了这些值,所以也许枚举是有序的？

enum BRANCH {
  A,B, ... Y,Z;
}

然后在你的代码中:

BRANCH branch = BRANCH.valueOf( WORD[ INDEX ] );

此外,"A" == "A"根据"A"的对象标识,您的代码中可能存在错误,可能是错误的.